A Hybrid Model Of Gene Expression Profiles Reducing Based On The Complex Use Of Fuzzy Inference System And Clustering Quality Criteria

The paper presents the hybrid model of gene expression profiles reducing based on complex use of fuzzy inference system and clustering quality criteria. The average of absolute values, the variance and Shannon entropy of the gene expression profiles were used as the input parameters of fuzzy inference system. The quality of gene expression profiles was used as output parameter of the model. The boundary value of the output parameter of the model was determined based on the minimum value of the clustering quality criteria which takes into account both the character of the objects distribution within clusters relative to the mass centers of the clusters, where these objects are, and the character of the clusters mass centers distribution in the feature space. Practical implementation of the proposed model allows us to divide the gene expression profiles into informative and non-informative objectively for purpose of the further investigation of the character of genes interconnection in the studied object.